BACKGROUND: The prevalence of atrial fibrillation (AF) has been increasing in hemodialysis patients (HD). Despite guideline recommendations and the increased risk of stroke, use of warfarin for anticoagulation has been low. Little is known about treatment uptake and thromboembolic risk management in these patients. International normalized ratio (INR) is a better measure of warfarin use, adherence and risk management than pharmacy claims data alone.
OBJECTIVES: To describe anticoagulation treatment and risk management for stroke prophylaxis in a cohort of HD patients with AF.
METHODS: We conducted a retrospective cohort study of HD patients with AF newly initiating warfarin using linked administrative data from the US Renal Data System and clinical data from a large US dialysis provider (2007–2011). Adult patients with continuous Medicare part A, B and D coverage and no warfarin or dabigatran use in a 6-month baseline period prior to incident AF diagnosis were included. Patients were followed from warfarin initiation (index date) to end of study period, death or administrative censoring. Individual INR values were categorized as low (<2.0), target (2.0–3.0) or high (>3.0). Monthly INR categories were estimated using all available individual lab draws in a given month and classified as a) Low (at least 1 low – none high); b) target (all in range); c) high (at least 1 high – none low); d) not in range (at least 1 low and 1 high). Thromboembolic risk management was described over time as proportion of patients in each of the monthly INR categories.
RESULTS: Of the 17,211 anti-coagulant naïve patients with incident AF, 3798 patients (22%) were warfarin initiators. INRs were available for 1,970 (52%) of warfarin initiators. Using an ITT approach not censoring at discontinuation, monthly INRs were in target range for about 8%, not in range for 15%, low for 70%, and high for less than 10% of patients over the follow-up period. However, compared to baseline, monthly INRs following immediately after warfarin initiation demonstrated a 10% and 5% initial decrease in number of patients with low and high monthly INRs respectively.
CONCLUSIONS: Warfarin use in HD population with AF continues to be very low with some evidence of off-label dabigatran use. Risk of ischemic stroke remains high in three-quarters of warfarin initiators due to inadequate anticoagulation. Concerns regarding bleeding risk may be overestimated. Further assessment of confounding, censoring and vascular outcomes will be conducted.
